Problem: Following along with a printed recipe is difficult, and following an online video is hard as well
Solution: Recipes broken down into small video clips that my brain can handle
Product: An iOS app that lets you find recipes by chef. Then you can flip through each step of the recipe in short 5-10 second video clips. Chefs can use it as a social network to promote their brand, or to sell recipes directly.
Skills Learned: Swift. Mobile video. Amazon mobile backend. DynamoDB.
Problem: Attention spans are shrinking and younger people don't want to read news articles, or watch 3 minute videos
Solution: A daily digest of the news broken down into 5 second clips
Product: An iOS app that once a day sends you the top 5 news stories. Someone (me) summarizes the news story into 5 second clips that you can swipe between.
Skills Learned: Content creation (me reading the news). Swift. Mobile video. Amazon mobile backend. DynamoDB. Swipe to interact.
Problem: My shoes were always dirty and smelly. How could I quickly get them cleaned?
Solution: An on demand shoe cleaning company
Product: a quick signup page with Stripe integration. Site was picked up from Reddit. Within minutes it was blogged about on a variety of blogs. I personally delivered and cleaned 50 pairs of shoes before finding out that I didn't want to clean shoes for a living, despite there being huge demand to do so!
Skills Learned:Value of viral marketing. Product where demand far outstripped my supply. How to optimize deliveries around NYC. How to clean shoes. Quick splash pages with Stripe integration.
Problem: It's time consuming to identify recipes you want to cook, compile a grocery list, cross reference with what you already have, and then shop for the ingredients
Solution: Recipe curation service where you can automatically have the ingredients delivered
Product: A web app where you can swipe through our pre-compiled recipes. Select the ones you like and we automatically build the shopping list. De-select ingredients you already have and with a single click have them delivered (by me!)
Skills Learned: Mobile friendly web app. Stripe integration. Hustle (building recipe ingredient lists manually).
Problem: I was watching Wedding Crashers at the gym on cable and thought to myself "How many other people are watching this right now? This movie's hilarious I'd love to chat with them"
Solution: Anonymous mobile chats built around actual live TV shows or movies, or movies everyone plans to watch at the same time
Product: An iOS and Android app called Gabba. You find a TV show that's playing currently and join the conversation directly around it. Completely anonymous and zero logins required. Utilized a Parse backend for both iOS and Android.
Skills Learned:Swift. Android. Parse (RIP). Mobile notifications. Biggest learning is that it's very tough to grow a chat community from scratch. You have to be willing to play 'sock-puppets' for awhile and eventually I got bored.
Problem: Meetings always run way too long, or get off track
Solution: A timekeeper for each agenda item of the meeting
Product: Standy, a web app built on Meteor in Javascript where you can instantly create a meeting, add agenda items, and set a time limit. You would then put this up on the screen during the meeting, allowing you to take notes and assign action items live. A countdown timer during the meeting keeps everyone in check to ensure all topics are covered.
Problem: I have 2k photos on my phone, and zero actually printed out because it's annoying to do so.
Solution: Swipe through photos on your phone, automatically print the ones you like
Product: Swipe and Print, an iOS app built in Objective-C. You can swipe through each photo Tinder-style, then automatically have them printed at your local Walgreens or Duane Reade. This required integrating with the Walgreens API, and I still receive a commission for every photo printed using this app
Skills Learned:Integrating with Walgreens API. Time required if API isn't public. Value in a cash app that continutally brings in revenue. And I still use it for printing my photos!
Problem: Donating to charities and social causes takes too much time and reseach
Solution: With two clicks, donate to a leading foundation for a cause you support.
Product: Donatr, an iOS app built in Objective-C where you can find 5 major causes (gun control, abortion rights, climate change), and select if you are "For" or "Against". We automatically find the top foundation aligned with your cause, and with Stripe integration enable you to donate directly to them. After building this and sending to the App store for review I found out that Apple doesn't allow you to make charitable donations through any app on their app store.
Skills Learned: READ THE APPLE REQUIREMENTS BEFORE BUILDING! I wish I had known that apple prevents donations. Seems silly right? That's a ton of money they are preventing going to good causes. Their suggestion FYI is to send people to the donation page site, which is a horrible user experience. I did learn how to integrate stripe into an app and about simple UX design.
Problem: We were searching for an application for our tech at Pairprep and wanted to learn more about risks that top educational companies were facing. Digging through 10-k annual reports for top risks facing companies is time consuming
Solution: A searchable database of annual report risks
Product: A linux bash script that crawled all 10-k reports for the last year and using heuristics copies the text from the risk section of the annual report. Then a Meteor app that indexes that data into a MongoDB database. And finally, a thin UI for searching through all of those risks.
Skills Learned: Web scraping. Bash. Search indexing. Search UI.
The Duffalo - cheap high quality weekend duffel bag
Problem: I've needed a good weekend bag for awhile, but all the good ones cost $300+
Solution: Importing bags from China and building a brand.
Product: The Duffalo, an old-school gym bag that costs $5 to make, and sells for $40. I just built a quick splash page for this one and posted a few ads on Facebook. Learned a ton about how to import from China.
Skills Learned: Chinese importing. Alibaba. Brand building is way way harder than you'd think.
Problem: I wasn't getting enough sleep, and losing an hour of sleep a day cuts nearly a decade off your life.
Solution: A sleep pack that can be given as a gift that has everything you need to get an extra hour of sleep a day.
Product: A branded box with memory foam pillow, a great eye mask, incredible ear plugs, and melatonin. All to be imported from China. Created a splash page again for this and integrated Stripe.
Problem: Working in the Columbia Startup Lab I kept overhearing non-technical founders saying they wished they could just talk with an impartial 3rd party CTO for an hour so they could validate their dev-team's work.
Solution: Hourly consultations with CTOs
Product: Just a simple page to schedule a time for you to get an hourly consultation with a CTO (me!)
Problem: I needed to contact teachers in the US to try and market our product for pairprep
Solution: Collect as many teacher email addresses as possible without paying anything.
Product: I built a web crawling scraper in Bash that started from an initial list and crawled the web for school web pages. Then did some heuristics to find the contact pages, and finally scraped all of the email addresses. An hour of scraping returned 70k email addresses, or roughly 2% of all teachers in the US. Apologies for the spam!
Skills Learned:Bash. Web scraping where the crawler continues searching outwards (as opposed to just going down a list). Heuristics for spotting emails. Growth hacking
Problem: I needed to find the physical address of every school in NYC for sending mailings for Pairprep
Solution: Google has already done the legwork, so just needed a bot that could pretend like it was a human to trick Google into returning results.
Product: A web app that started at a certain Lat/Lon location, and progressively moved around the map, returning all the schools nearby. The data comes from the actual Google Maps product, so needed to put in random wait times to convince it that I wasn't a bot.
Skills Learned:Web scraper that pretends like it's human. Adds random wait times. Coded in Bash.